R/lb_test.R

Defines functions print.lbtest lb.test

Documented in lb.test print.lbtest

lb.test <- function(obj, ...){
  if(is.zoo(obj))  obj <- as.matrix(obj)
  if(is.null(dim(obj))) stop("obj is a vector. Use `Box.test()'.")
  ndim <- ncol(obj)
  lags <- c(1, seq(5, 50, 5))
  nlag <- length(lags)
  
  #     stat <- matrix(0, nlag, ndim)
  #     p.val <- matrix(0, nlag, ndim)
  stat <- matrix(0, nlag, 2*ndim)    
  for(i in 1:ndim){
    for(j in 1:nlag){
      LB.test <- Box.test(obj[,i], lag = lags[j], type = "Ljung")
      stat[j, (2*i-1)] <- LB.test$statistic
      stat[j, 2*i] <- LB.test$p.value
      #             stat[j, i] <- LB.test$statistic
      #             p.val[j, i] <- LB.test$p.value
    }
  }
  
  odd.ind <- seq(1, 2*ndim, by = 2)
  evn.ind <- seq(2, 2*ndim, by = 2)
  Snames <- numeric(2*ndim)
  if(is.null(colnames(obj))){
    Snames[odd.ind] <- paste("Series", 1:ndim, sep=" ")
  } else {
    Snames[odd.ind] <- colnames(obj)
  }
  
  Snames[evn.ind] <- "p-value"
  colnames(stat) <- Snames
  # colnames(p.val) <- rep("p-value", ndim)
  rownames(stat) <- paste("Lag", lags)
  # rownames(p.val) <- paste("Lag", lags)
  
  #ans <- list(stat = stat, p.val = p.val)    
  
  class(stat) <- "lbtest"
  return(stat)
}

print.lbtest <- function(x, digits = 3, ...){
  cat("\nLjung-Box Test for Serial Correlations", "\n")
  print.default(round(unclass(x), digits = digits), print.gap = 1, quote = FALSE)
  
}

Try the ccgarch2 package in your browser

Any scripts or data that you put into this service are public.

ccgarch2 documentation built on May 31, 2017, 4:23 a.m.